Kinetics of the decomposition of total aliphatic waxes in olive oil

Kinetics of the Decomposition of Total Aliphatic Waxes
in Olive Oil During Deodorization
R.M. Tubaileha, Mª.M. Graciani Constanteb, M. León Camachoa,
A. López Lópeza,*, and E. Graciani Constantea
a
Instituto de la Grasa Consejo Superior de Investigaciones Cientificas (CSIC), 41012 Sevilla, Spain, and
Departamento de Química Física, Facultad de Química, Universidad de Sevilla, 41012, Sevilla, Spain
b
ABSTRACT: Changes in the content of aliphatic waxes during
industrial deodorization and/or physical refining of bleached
olive oil were studied in an experimental discontinuous pilot
plant of 250 kg deodorizer using nitrogen as stripping gas in place
of steam. The kinetic constants for the decomposition of waxes
during the deodorization process were determined. The reaction
orders studied are zero (or can be considered zero) within the
working interval. The values of rate constants, activation energy,
frequency factor, increment of activation Gibbs free energy, activation enthalpy, and activation entropy are established.
Paper no. J10141 in JAOCS 79, 971–976 (October 2002).
KEY WORDS:
Kinetic, olive oil, physical refining, wax.
During olive oil deodorization (in both chemical and physical
processes) there is a change in the content of aliphatic waxes
with respect to the bleached oil. When the bleached olive oil
reaches the deodorization temperature, the content of aliphatic
waxes is maximal (generally, all the possible waxes that can
be formed from its content in fatty alcohols). However,
throughout the deodorization process there is a diminution of
the wax content that is due to the decomposition of waxes and
the distillation of the alcohols formed during this process and,
in the case of physical refining, the distillation of the FFA present in the oil (1).
The formation–decomposition of waxes during olive oil
deodorization is a reaction that takes place at high temperatures (between 180 and 270°C, at maximum) and under vacuum (in general, between 3 and 10 torr of absolute pressure).
It is an esterification–deesterification reaction between one
FA and a fatty alcohol that can be expressed by
→ wax (aliphatic, which
FA + fatty alcohol ←
can be quantitatively determined) + water
[1]
The reaction also can be described as a transesterification
reaction between a fatty alcohol and an acylglycerol, and its
two phases can be described as
→ glycerol, MAG, or DAG + FA [2]
MAG, DAG, or TAG + water ←
*To whom correspondence should be addressed at Instituto de la Grasa
CSIC, Avda. Padre García Tejèro 4, 41012 Sevilla, Spain. E-mail: [email protected]
Copyright © 2002 by AOCS Press
In general, waxes are formed mainly from the FFA. These FA
then react as indicated in Equation 1 (second phase of this
transesterification reaction).
There are reports in the scientific literature on esterification reactions between FA and short-chain alcohols, because
they are used for the qualitative and quantitative determination of such acids as methyl esters by GC. The direct esterification is catalyzed by strong acids and is usually achieved by
sulfuric acid/methanol and/or hydrochloric acid/methanol.
The transesterification of acylglycerols is catalyzed by strong
alkalis, usually alcoholic KOH or sodium methylate. In this
kind of reaction the catalyst has a great influence on the reaction rate (2).
Kinetic studies on the esterification reaction between alcohols and organic acids have been performed using short-chain
alcohols and short-chain FA (in general, the monosodiumsubstituted salts of diacids), because they allow one to do the
studies with aqueous or MeOH/H2O solutions. Examples of
such researches are the studies performed to determine the effect of salts on the kinetics of formation of monoesters of
ethyl malonate, methyl glutarate, and ethyl adipate performed
by Sánchez Burgos et al. (3,4). These reactions were studied
for the first time by Nielsen (5) in 1936 and a previous mention of them can be found in the work of Meyer (6).
The references found in the literature are not relevant for
describing the formation–decomposition of aliphatic waxes
during the process of deodorizing olive oil. In this case, the
substances involved in the reaction are high M.W. alcohols—
linear molecules of 18 (or more) carbon atoms—and FA. The
reaction takes place in olive oil (which is not an alcoholic,
aqueous, or water–alcohol medium), without the use of any
catalyst (strong acid or alkali), at temperatures between 180
and 270°C and under vacuum.
With respect to the possibility of changes in the acylglycerol structures, the assays performed by Gracian and Mancha
(7) showed clearly that physical refining produces only small
changes that, for most edible oils and fats, are of little significance. However, in the case of the olive oil, such changes
have great importance because of the strict regulation in relation to their contents (8–10).
These studies cannot be considered as models for the formation of waxes, since they are related only to the transesterification of acylglycerols.
971
JAOCS, Vol. 79, no. 10 (2002)
972
R.M. TUBAILEH ET AL.
The aim of this work was to study the kinetics of decomposition of waxes during the process of deodorizing olive oil.
These parameters are of interest for the olive oil refining industries, which want to develop more efficient processes that
lead to improved quality of the treated oils, where limits for
wax content are established by European Common Market
(EEC) regulations (8,9). In addition, refined oils must contain
no more than a minimal proportion of waxes so as to avoid
rejection by consumers owing to the turbidity produced by
the crystallization of the excess wax. Removal of waxes by
crystallization and/or filtration is an expensive process because
of the neutral fat losses and the energy necessary to achieve
the separation.
EXPERIMENTAL PROCEDURES
Physical refining. Olive oil with a FFA content of 1.85% was
bleached in the pilot plant of the Instituto de la Grasa according to the technique described in Reference 1 (0.5% clayType OD, Minas de Gador, Almería, Spain; 90°C, 30 min, absolute pressure lower than 30 mm Hg). Batches were taken
from a homogeneous mixture of bleached oil. Bleached olive
oil was physically refined in a discontinuous deodorizer of
250 kg maximum capacity. The technique and the equipment
were described (1,10). The experiment carried out for olive
oil described in Reference 1 was planned according to a centered cube 23 experimental design (complete factorial design)
with three central points. The experimental treatments (Table
1, treatments E-1 to E-11) were performed in random order,
and the range of values for each variable was selected according to the optimal values from a previous similar experiment.
In each experiment the N2 flow was introduced into the deodorizer, which was already charged with the corresponding
amount of bleached oil. The temperature was raised to 100°C,
to ensure heat transfer and protection of the oil, before the refining began. The rate of N2 flow (±0.04 m3/h), the operating
temperature (±0.5°C), and the oil height in the deodorizer
(measured by weight ± 0.5 kg) (1) were controlled according
to the design levels (Table 1). Time (± 1 min) was computed
from the point at which the oil reached the required temperature. Each batch was sampled at 0, 1.5, 3, 4, and 5 h. The N2
flow was shut down after the batch cooled to 100°C.
Chemical refining. Lampante olive oil (3.9% FFA content)
was heated at 40°C for 20 min with 0.2% of commercial
phosphoric acid (85%, w/w), and the excess acid was neutralized with sodium hydroxide at the same temperature. After
20 min, the temperature was raised to 90°C to speed soap formation, and the soap was separated by centrifugation 20 min
later.
The neutralized oil was washed three times with water and
dried. Then it was bleached under the same conditions mentioned above. The deodorization process was performed
under the same conditions as in the physical refining but at
180°C. Samples were taken when the oil reached 180°C (t =
0) and after 1, 2, 3, and 4 h of processing.
JAOCS, Vol. 79, no. 10 (2002)
TABLE 1
Experimental Design, Showing the Values of the Variables Studied
and Conditions of the Different Deodorization Assays Performed
Flow of N2a
3
[m (n.c.)t
1.5
−1
Oil weight (kg)
−1
h ]
240°C
2.0
2.5
250°C
125 (E-1)b
157 (E-3)
260°C
125 (E-2)
175 (E-4)
150 (E-5)
150 (E-6)
150 (E-7)
125 (E-8)
125 (E-9)
125 (E-10)
125 (E-11)
a
Notes: n.c. = normal conditions; t = tonnes.
In parentheses are the treatment codes.
b
Determination of wax content. Wax content was determined according to the EEC regulations for olive oil and pomace oil, Regulations (CEE) no. 2568/91 EEC (8) modified by
Regulations (CEE) no. 183/93 (9).
The olive oil sample was prepared with 0.5 mg of oil and
1 mL of internal standard (0.1 mg/mL lauryl arachidate, C32)
with a few drops of Sudan I as colorant; because the oil was
stored in the freezer, it was necessary to heat it about 30 min
at 40–50°C, with vigorous stirring, before taking the sample.
Waxes were separated on a silica gel column using
hexane/ether (98.5:1.5) as eluent. The first fraction, with a polarity lower than TAG, was concentrated to dryness and then
recovered with 4 mL of n-heptane. GC analysis was performed by injecting 1 µL of this solution onto a fused-silica
capillary column (15 m length, 0.25 mm i.d., and 0.10 µm
thickness) with an on-column injector. The chromatographic
conditions were: HP chromatograph (model 5890-II) with automatic injector and FID. The temperature program was from
80 to 120°C, at about 30°C/min, and from 120 to 340°C at
about 5°C/min. The temperatures of the detector and the injector were 320°C. Hydrogen flow was 25–35 cm3/min and
analysis time was 40 min (1).
Analytical error was calculated from the 13 samples of oils
with different amounts of analyte. Each sample was divided
into two subsamples, and each subsample was injected in duplicate into the gas chromatograph. The error in sample determination was evaluated as ±4% (1).
Kinetic study. To estimate the reaction order, data were adjusted to reactions of order 0, 1, 2, and 3 by the method of
least squares. Then residual variances and correlation coefficients were compared. Equations were fitted considering two
different situations: (i) Each deodorization assayed was independent; (ii) assays at the same temperature are replications
made under similar conditions as the other two independent
variables (since these variables have a limited influence on
the content of waxes in comparison with the deodorization
temperature) (1).
This latter hypothesis implies the inclusion of the variances derived from the other two variables in the total variance of the fitted equation. This implies that, if the equations
are fitted independently for each deodorization assay
DECOMPOSITION KINETICS OF ALIPHATIC WAXES IN OLIVE OIL
performed at the same temperature, the fitted curve may be
the same, but the confidence intervals corresponding to the
experimental points will vary their positions with respect to
the curve and, in consequence, the integration constant also
will change. To fit the equation in these conditions it is
enough to establish that the curves obtained at the same temperature are parallel (i.e., they have the same slope but different constants for x = 0). The common slope is thus an estimation of the value of k, expressed as ppm h−1. Once the values
of k are known, the other thermodynamic constants can be
easily estimated.
RESULTS AND DISCUSSION
The kinetics of the decomposition of aliphatic waxes can be
quantified during the deodorization process and the parameters estimated. These results are useful in making an estimation of the content of waxes in refined olive oils.
The reaction rate (v) at a specific time is, according to the
law of mass action, directly proportional to the product of the
concentrations of the respective substances (ri) involved in
the reaction (cri). The power to which each concentration is
raised is generally an integer (ai). The constant of proportionality (k) is known as the “specific rate” or “rate constant.”
973
ln k = lnA − E ≠/RT
By estimating the rate constant (k) at different temperatures
(T), the values of A and E ≠ can be calculated with Equation 6,
ln (cri) = − E ≠/RT + ln(t) + ln(A)
[6]
This expression is often written as:
ln (cri) = −E ≠/RT + B · ln(t) + ln(A)
[7]
where B (fitting coefficient) must be 1 when the reactions are,
or can be considered as, order 0 in the interval of time studied.
For a simple reaction such as wax formation, and since the
reaction occurs at constant temperature and pressure throughout the deodorization period, the values of ∆G ≠, ∆S ≠, and ∆H ≠
can be calculated or estimated from the values found for the
rate constant (12):
−∆G≠ = [ln k − ln (kB · T/h)]· RT
[8]
and by the equation (13):
∆G ≠ = ∆H≠ − T·∆S≠
v = k ∏ criai
[5]
[9]
[3]
By definition, the order of a reaction is the sum of the powers that appear in Equation 3. At the same time, each power
(ai) is known as a partial order (with respect to the substance,
ri). Several mathematical and/or graphical procedures are
available to determine the order, and the order of the reaction
adopted is the one that gives the lowest variance and the highest correlation coefficient (11).
The procedures are summarized in Table 2.
Except for the odd exception, the reaction rate increases
with temperature. Arrhenius established this dependence (12,
13),
k = A · exp (−E≠/RT)
where ∆G≠ = increment of activation Gibbs free energy; T =
absolute temperature; R = 1.987 cal · K−1 · mol−1; kB = Boltzmann constant = 1.38026 · 10−16 erg · K−1; h = Planck constant
= 6.6252 · 10−27 erg · s; k = rate constant; ∆H ≠ = increment of
activation enthalpy; ∆S≠ = increment of activation entropy.
The values of the contents of waxes in the different assays
and samples analyzed are shown in Figures 1–4.
[4]
in which the constant A is termed the “frequency factor” or
“pre-exponential factor” and E≠ is the “activation energy.”
Expressing this equation logarithmically,
TABLE 2
Equations for the Estimation of the Kinetic Parameters
According to the Reaction Ordera
Order
Differentiation method
0
1
2
3
dcri /dt = k
dcri /dt = k · (a − cri )
dcri /dt = k · (a − cri )2
dcri /dt = k · (a − cri )3
a
Integration method
k · t = cri
k · t = ln[a/(a − cri )]
k · t = [1/(a − cri )] − 1/a
k · t = [1/(3 − 1)]·[(1/(a − cri )2] − (1/a2)]
Volume: V; concentration: cri = molsri /V; concentration: cri = ppmri =
(molsri . Pmp)/(V·pe) ; M.W. (weighted): Pmp; specific weight of the oil:
pe ; t: time; k: rate constant; a: constant.
FIG. 1. Contents of total waxes during the refining assays performed at
240°C. Data were fitted to each assay independently, considering a reaction order of zero. Significance was determined by F test of the variance analysis of the regression. For treatment codes see Table 1.
JAOCS, Vol. 79, no. 10 (2002)
974
R.M. TUBAILEH ET AL.
FIG. 2. Contents of total waxes during the refining assays performed at
250°C. Data were fitted to each assay independently, considering a reaction order of zero. For treatment codes see Table 1.
Because of the lack of pertinent prior information the possibility that the order of the reaction is 0, 1, 2, and 3 has been
checked. Bearing in mind that the most appropriate order is
the one that produces the highest correlation coefficient and
lowest variance, the order of reaction for our data can be established as 0. Then the order of the reaction is, within the interval considered, 0.
Figures 1–4 show the fitted curves for each of the assays.
It is evident that the diverse curves show similar slopes, confirming the hypothesis that they can be consider as parallel
for the calculation of k.
Table 3 shows the estimated values of ∆G≠ and k, assuming that the assays performed at the same temperature may be
considered as replications.
FIG. 3. Contents of total waxes during the refining assays performed at
260°C. Data were fitted to each assay independently, considering a reaction order of zero. For treatment codes see Table 1.
JAOCS, Vol. 79, no. 10 (2002)
FIG. 4. Contents of total waxes during the refining assays performed at
180°C. Data were fitted to each wax independently, considering a reaction order of zero.
To calculate these data, one must express the k of the reaction in mols/L·s, using 0.826, 0.797, 0.793, and 0.788 g/mL
as olive oil density at 180, 240, 250, and 260°C, respectively.
The results suggest that the lower M.W. waxes are more
easily decomposed during chemical refining (higher k values)
(Table 3). This means that by applying chemical refining, the
refined oil will have higher proportion of high M.W. waxes.
However, in physical refining, there is no consistent tendency
with respect to the M.W. of waxes. Apparently, the higher
temperature of the process counteracts the effect of M.W. in
the decomposition reaction and produces a similar rate across
the M.W. range studied.
∆G≠ (Table 3) increases with temperature in all waxes during physical refining, and its values are independent of the
M.W. of waxes. In chemical refining ∆G≠ are lower corresponding to a lower processing temperature.
Estimated values of E ≠, A, ∆H ≠, and ∆S≠, for the values of
k found, are shown in Table 4. This table shows that values
for ∆S≠ are similar in the range of wax M.W. studied. In contrast, ∆H≠ increases with wax M.W. This means that, in the
temperature interval studied, the energy to be supplied will
depend on the wax M.W.
This work has permitted the estimation of different thermodynamic parameters of wax decomposition during the deodorization of olive oil. For the first time, values of E ≠, A, and
increments of ∆G≠, ∆H≠, and ∆S≠ have been established.
A good knowledge of wax decomposition kinetics will
permit optimization of the processing variables to produce the
highest wax decomposition and will avoid any further elimination of wax by crystallization from the deodorized oils before commercialization. Furthermore, elimination of waxes
during physical refining results in lower neutral fat losses and
lower energy consumption because it will not be necessary to
separate them by crystallization.
DECOMPOSITION KINETICS OF ALIPHATIC WAXES IN OLIVE OIL
975
TABLE 3
Order (according to the hypothesis established) Within the Interval of Temperatures Studied
for the Formation–Decomposition of the Aliphatic Waxes Quantified in the Olive Oil, k, and ∆G ≠
Wax
C-36
C-38
C-40
C-42
C-44
C-46
Totalsa
a
Type of
refining
Temperature
(°C)
Order of
reaction
k
(ppm h−1)
k
(mol L−1 s−1)
∆G ≠
(kcal mol−1)
240
250
260
180
240
250
260
180
240
250
260
180
240
250
260
180
240
250
260
180
240
250
260
180
240
250
260
180
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8.7
8.4
7.7
26.9
11.9
11.1
9.2
7.9
17.0
15.9
14.2
20.9
16.0
16.1
14.9
8.4
12.5
11.9
12.2
3.8
6.6
7.4
6.5
1.1
72.8
70.9
64.8
68.8
5.7 · 10−9
5.5 · 10−9
5.1 · 10−9
1.7 · 10−8
7.4 · 10−9
6.9 · 10−9
5.5 · 10−9
4.7 · 10−9
1.0 · 10−8
9.5 · 10−9
8.5 · 10−9
1.2 · 10−9
9.1 · 10−9
9.2 · 10−9
8.5 · 10−9
4.6 · 10−9
6.8 · 10−9
6.5 . 10−9
6.7 · 10−9
2.0 · 10−9
3.4 · 10−9
3.9 · 10−9
3.4 · 10−9
5.5 · 10−10
4.2 . 10−8
4.1 · 10−8
3.8 . 10−8
3.8 · 10−8
49.8
50.8
52.0
42.8
49.5
50.5
52.0
44.0
49.2
50.2
51.5
43.2
49.3
50.2
51.5
44.0
49.6
50.6
51.7
44.8
50.3
51.1
52.5
45.9
47.7
48.7
49.9
42.1
Physical
Physical
Physical
Chemical
Physical
Physical
Physical
Chemical
Physical
Physical
Physical
Chemical
Physical
Physical
Physical
Chemical
Physical
Physical
Physical
Chemical
Physical
Physical
Physical
Chemical
Physical
Physical
Physical
Chemical
M.W. (weighted) = 602.
TABLE 4
Estimated Values of E ≠, A, ∆H≠, and ∆S ≠ in Physical Refining (240–260°C)
Waxes
E≠
(kcal mol−1)
A
(mol L−1 h−1)
C-36
C-38
C-40
C-42
C-44
C-46
Totals
−3.014
−8.033
−4.404
−1.837
−0.415
0.048
−2.709
2.98 · 10−10
2.87 · 10−12
1.34 · 10−10
1.52 · 10−9
4.47 · 10−9
3.73 · 10−9
2.97 · 10−9
ACKNOWLEDGMENTS
The authors express their thanks to Esperanza Oliveras Moreno, Enrique Pinto García, and Manuel Serrano García, staff members of
the Instituto de la Grasa, for the labor and assistance carried out; to
Aceites Toledo oil companies for supplying the oils; to Comisión
Interministerial de Ciencia y Tecnologia (CICYT) for their financial
support (ALI 91-0720, PTR 93-0061, ALI 94-0716), and to Carburos Metálicos S.A. for their collaboration.
REFERENCES
1. Tubaileh, R.M., A. Garrido Fernández, M. León Camacho, M.V.
Ruiz Méndez, and E. Graciani Constante, Effects of Deodorization
Variables (physical refining) in Wax and Fatty Alcohols Content of
Refined Olive Oil, J. Am. Oil Chem. Soc. 79:101–104 (2002).
∆H ≠
(kcal mol−1)
−8.8
−13.8
−10.1
−7.6
−6.1
−5.9
−8.3
∆S ≠
(kcal mol−1)
−0.11
−0.12
−0.12
−0.11
−0.11
−0.11
−0.11
2. Noureddini, H., and D. Zhu, Kinetics of Transesterification of
Soybean Oil, Ibid. 74:1457–1463 (1997).
3. Maraver, J., J.I. Tejera, J. Rodríguez Velasco, F. Sánchez Burgos, and E. Dobarganes, Efectos salinos en la hidrólisis del
semiester etílico del ácido malónico en medio básico, An. Quim.
79:543–544 (1983).
4. Martín Herrera, M., J.J. Maraver Puig, and F. Sánchez Burgos,
Kinetics of Salt Effects in the Hydrolysis of Ethyl Malonate,
Methyl Glutarate and Ethyl Adipate Semiesters, Collec. Czech.
Chem. Commun. 51:2781–2785 (1986).
5. Nielsen, R.F., Kinetic Salt Effect in Saponification of Ester Ions,
J. Am. Chem. Soc. 58:206–215 (1936).
6. Meyer, J., Saponfication in Stages of Esters of Dibasic Acids II,
Z. Physik. Chem. 67:278–308 (1909).
7. Gracian, J., and M. Mancha, Fenómenos de esterificación
durante el calentamiento de aceites de oliva y de orujo y su inJAOCS, Vol. 79, no. 10 (2002)
976
R.M. TUBAILEH ET AL.
fluencia en la alteración de su estructura glicerídica, Grasas
Aceites 22:22–29 (1971).
8. Reglamento (CEE) no. 2568/91 de la Comisión de 11/06/91 relativo a las características de los aceites de oliva y de los aceites
de orujo de oliva y sobre sus métodos de análisis, Diario Oficial
de las Comunidades Europeas de 5/09/91 (1991) No. L
248/1–248/83.
9. Reglamento (CEE) no. 183/93 de la Comisión de 29/01/93 por
el que se modifica el Reglamento (CEE) no. 2568/91 de la
Comisión de 11/06/91 relativo a las características de los aceites
de oliva y de los aceites de orujo de oliva y sobre sus métodos
de análisis, Diario Oficial de las Comunidades Europeas de
30/01/93 (1993) No. L 22/58–22/68.
JAOCS, Vol. 79, no. 10 (2002)
10. León Camacho, M., M.V. Ruiz Méndez, and E. Graciani Constante, Changes in Olive Oil Components During Deodorization
and/or Physical Refining at the Pilot Plant Scale Using Nitrogen
as Stripping Gas, Fett/Lipid 101:38–43 (1999).
11. Rodríguez Velasco, J., F. Sánchez Burgos, and M. Domínguez
Pérez, Lecciones de cinética química, Publicaciones de la Universidad de Sevilla, 1975, pp. 11–25.
12. Pannetier, G. and P. Souchay, Chimie Générale. Cinétique
Chimique, Masson et Cie Editeurs, Paris, 1964, pp. 87–96.
13. Castellan, G.W., Fisicoquímica, 2ª Edición, Addison-Wesley
Interamericano, Bogotá, Brasil, 1998, pp. 216–234.
[Received October 31, 2001; accepted June 11, 2002]